Scan Visual — 3 Options

pick one and I'll integrate it · same design tokens as the rest of the site

Option A Tight Terminal
Same terminal, 3× faster, no ASCII borders

Keeps the authentic CLI feel. Cuts the box-drawing characters and pip install preamble. Leads with color-coded risk rows immediately — chain alert appears after ~2s. Loops automatically.

agentward scan ~/

Option B Split Panel
Terminal log on the left, live risk panel on the right

Left side shows condensed discovery log. Right side fills in a risk card for each tool as it's found. The visitor can follow the story on the left or just read the results on the right — two reading modes, same data.

Risk findings
⚠ Skill chain detected
email-mgr → web-browser
lateral data movement risk
3 critical 4 high 1 medium 1 low

Option C Risk Map
Static card grid — instant comprehension, no waiting

No animation. Every tool visible at once, color-coded by risk level, with chain connections shown inline. Visitor understands the full picture in under 2 seconds. Works perfectly on mobile.

scan output — 9 tools, 3 sources $ agentward scan ~/
MCP
filesystem
read · write · delete
🔴 CRITICAL
Skill
email-mgr
send · read · delete
⚠ HIGH
Skill
web-browser
navigate · js · download
⚠ HIGH
SDK
shell_exec
subprocess · env
🔴 CRITICAL
Skill
finance
read · network
🔴 CRITICAL
MCP
slack
all-channels
⚠ HIGH
SDK
send_email
smtp · attachments
⚠ HIGH
MCP
postgres
read · write
◆ MEDIUM
MCP
github
repos · issues
✓ LOW
Skill chain: email-mgr → web-browser
Lateral movement — email content could exfiltrate via browsing
AgentWard — a product from OpenSafe Inc